List view

• M4: Hardening & release

  Docs, reproducibility, calibration-in-CI, observability (tracking)

  No due date

  •0/1 issues closed
  0% complete1 open0 closed

• M3: Public website

  Dual-metric public leaderboard, certificate drilldown, trajectory, CI publish

  No due date

  •0/2 issues closed
  0% complete2 open0 closed

• M2: Fair multi-model + trustworthy judgment

  Scheduler runs sessions across models fairly; hardened verifier + novelty

  No due date

  •0/2 issues closed
  0% complete2 open0 closed

• M1: MVP

  Working MVP — one model, real numbers end-to-end via pred CLI

  No due date

  •0/3 issues closed
  0% complete3 open0 closed